Engineering Cell Fate through Reprogramming and Differentiation
Our previous research in the field of cell fate reprogramming has demonstrated that CD34+ hematopoietic progenitor cells are amenable to rapid reprogramming (Loh et al., Blood 2009). In contrast, while terminally differentiated human T-cells can be reprogrammed, the process takes nearly a month (Loh et al., Cell Stem Cell 2010). Additionally, we showed that by targeting erythroblasts in circulation, it is possible to reprogram human blood cells using just a single drop of blood (Loh et al., Stem Cell Translational Medicine; IP: Loh YH, Tan TK, Method for inducing pluripotency in a hematopoietic cell, IMC/P/08345/04/US). We also developed a rapid method using lentiviruses, miniaturizing the process in a 384-well format and utilizing high-throughput imaging to assess the effects of genome-wide siRNA perturbation during the early stages of human reprogramming (Toh et al Cell Rep. 2016). Our screening identified functional repressors and effectors that hinder or facilitate the reprogramming process. Notably, a combinatorial knockdown of five repressors (SMAD3, ZMYM2, SFRS11, SAE1, and ESET) synergistically yielded approximately 85% TRA-1-60-positive cells (highest efficiency). Interestingly, ESET, ZMYM2, and SAE1 were also identified in our Yang et al. screen (Cell 2015) for ERV repression, suggesting that these factors may alter the epigenome-wide state of the parental cells, possibly making it more conducive to transitioning to new cell states during reprogramming. To explore this possibility further, we examined histone variant H3.3, known to suppress ERVs. In the study, we investigated the dynamic changes in H3.3 deposition during cellular reprogramming. H3.3 helps maintain the characteristics of parental cells during this process; its removal at early stages increases reprogramming efficiency. In contrast, H3.3 deposition on genes linked to the newly reprogrammed lineage is crucial, as its depletion in later phases can disrupt the process. Hence, our findings indicate the bimodal role of H3.3, as it is vital for preserving the initial parental fibroblastic cell fate while facilitating the transition to a new cell fate during the acquisition of pluripotency in reprogramming (Fang et al., Nat Commun. 2018).
Despite its widespread application, cellular reprogramming is often hindered by low efficiency. To better understand the heterogeneity and mechanisms underlying successful reprogramming, we utilized single-cell RNA sequencing (scRNA-Seq) and single-cell assay for transposase-accessible chromatin (scATAC-Seq) to profile reprogrammed cells at various time points. Our analysis revealed that reprogramming cells follow an asynchronous trajectory and give rise to diverse subpopulations. We also identified fluorescent probes and surface markers that enriched early reprogrammed human cells, and the combinatory use of these markers allowed for the precise segregation of early-intermediate cells with varying reprogramming potentials. Notably, we uncovered that the binary choice between a FOSL1-centric and TEAD4-centric regulatory network plays a crucial role in determining the success of reprogramming. By either knocking down FOSL1 or overexpressing TEAD4, we can significantly enhance reprogramming efficiency (Fig 3) (Xing et al., Sci Adv . 2020; IP: Live-cell Fluorescent Probe BDD2-C8 Detects Early Reprogramming Cells (2018) IMC/Z/10521).
Our previous work on cell differentiation has primarily focused on hematopoietic stem and progenitor cells (HSPCs) (Cheng et al., Nat Commun, 2016; Cipta et al., Nat Cell Biol. 2022) and myeloid lineage cells (Sivalingam et al., Tissue Eng Part C Methods, 2016; Sivalingam et al., Haematologica, 2018; Sivalingam, Stem Cell Reports, 2021). More recently, we developed an inducible system to directly reprogram mouse embryonic fibroblasts into HSPCs by expressing the transcription factors SCL and LMO2. The study identified two critical decision points during reprogramming: one associated with cell cycle state and another linked to competing fate decisions. Inhibiting pathways related to neuronal fates significantly improved reprogramming efficiency. This study highlights key reprogramming events and potential targets for enhancing efficiency (Shafiq et al., Stem Cell Reports, 2025). In collaboration with our immunologist colleagues, we are applying our expertise in stem cell differentiation to enhance the engineering of Natural Killer cells (funded by IAF-PP – PANAKEIA; and IAF-ICP – A*STAR/SCG Joint lab) and T cells for cancer immunotherapies (funded by NRF CRP – SPECTRA). Furthermore, due to our proficiency in reprogramming and engineering cell fates, we have established a portfolio of intellectual properties, some of which are licensed to companies such as Cytomed/Puricell (GMP-compatible reprogramming of human blood cells, IMC/Z/11729), Innocellular (cost-effective xeno-free medium for human mesenchymal stem cell expansion, IMC/Z/14328; chemically defined medium for human mesenchymal stem cell expansion, IMC/Z/2024-295; chemically defined medium for human pluripotent stem cell expansion, Z2024-465), and Genovn Therapeutics (highly efficient method for deriving genomic imprint-free clinical-grade human induced pluripotent stem cells, IMC/Z/13658).
Recently, our lab has developed an innovative technology for studying N6-methyladenosine (m6A) mRNA modifications, which play critical roles in the regulation of gene expression and are dynamically modulated across various cellular states. While bulk sequencing has facilitated global m6A profiling, there has been a significant lack of effective single-cell technologies to analyze m6A heterogeneity. To fill this gap, we present the single-nucleus m6A-CUT&Tag (sn-m6A-CT) method, which allows for the simultaneous profiling of m6A methylomes and transcriptomes within individual nuclei from mouse embryonic stem cells (mESCs). We have adapted m6A-CT for a droplet-based single-cell omics platform, showcasing its high-throughput capabilities for analyzing nuclei from thousands of cells across different cell types. Our findings demonstrate that sn-m6A-CT profiling accurately determines cell identities and generates cell-type-specific m6A methylome landscapes from heterogeneous populations. Using sn-m6A-CT, we successfully captured rare 2-cell populations within mESC cultures and identified previously unreported subsets of expanded pluripotent stem cells, characterized by high levels of m6A on Nodal and Sox2. This advancement not only enhances our understanding of epitranscriptomics but also provides critical insights into the dynamic landscape of cell identity and pluripotency (Hamashima et al., Mol Cell. 2023). We have further refined our single-nucleus m6A-CUT&Tag (sn-m6A-CT) method, evolving it into targeted-reverse transcription assisted RNA tagmentation with sequencing (scTART-seq). This advanced technique has enabled us to create a comprehensive single-cell epitranscriptomic map of mouse post-implantation development. Our analysis revealed lineage-specific transcription factors, unique cellular response peptides, membrane proteins, and growth factors exhibiting differential m6A enrichment associated with early embryonic lineages (Hamashima et al., In Preparation)
Other studies from our research employing single-cell and spatial platforms to study cell fate decisions has led to several significant discoveries. We identified progenitor-like cells in the iris compartment of the human retina (Gautam et al., Nat Commun. 2021), pinpointed geo-specific locations of rare mimetic thymic cells within the developing human thymus (Kamaraj et al., Nat Commun. 2025, Revision).